Video teleconferencing is a popular application for modern computing devices, allowing globally distributed individuals to work, learn, and socialize together. Compared to audio or text communication technologies, video communication may provide better opportunities for visual cues and other important human interactions. Typical video teleconferencing systems may identify which party to a teleconference is speaking based on audio levels of the transmitted signal. When displaying the parties, the speaker may be enlarged or otherwise emphasized, and other parties to the teleconference may be displayed uniformly or otherwise deemphasized. For teleconferences with a large number of participants such as presentations, each party other than the speaker may be difficult to view in detail or otherwise interact with.
Interestingness is a concept used in the field of data mining and knowledge discovery. Processing large amounts of data may involve applying data mining algorithms to identify patterns within the data and then focusing further analysis only on “interesting” patterns among those identified. Interestingness is thus a measure of which patterns identified within the data are interesting. In particular, an interestingness measure may indicate the relevance, accuracy, or representativeness of a particular pattern in relation to a data set. Various techniques and algorithms for measuring interestingness are surveyed in Sigal Sahar, Interestingness Measures: On Determining What Is Interesting, in Data Mining and Knowledge Discovery Handbook (Oded Maimon & Lior Rokach eds., 2d ed. 2010).